It's time for another GPU battle, though this one is a bit different from the norm considering the most recently released current generation GPUs from both AMD and Nvidia are by far their slowest: from AMD we have the Radeon RX 550 and on Nvidia's side is the GeForce GT 1030, a 384 CUDA core part that apparently isn't worthy of the GTX branding.
Priced well under the $110 GTX 1050, the GT 1030 costs only $70, which is even cheaper than the RX 550's MSRP of $80, though getting one at that price is nearly impossible at the moment. The mark up on the RX 550 isn't nearly as bad as the more higher-end models but right now you can expect to pay ~$90 and that makes it almost 30% more expensive than the GT 1030.
That said though, we have to wonder whether the Radeon is faster and if so by how much. If you have less than $100 to play with, which of these graphics cards should you buy? Complicating that question, these entry-level budget cards don't typically represent a great value as the slightly more expensive models generally deliver a much better price to performance ratio.
Our focus will be primarily on esports titles for this article, including games such as CS:GO, Overwatch and Dota 2. For testing I've used our Ryzen 3 test system with the 1300X installed. No overclock has been applied to the CPU but we are using 16GB of DDR4-3200 memory.
The card specifically in the green team's corner is MSI's GeForce GT 1030 2GH LP OC, while the red team will be represented by Asus' RX550-4G. Yes, the AMD graphics card has more VRAM but unfortunately this was all I was able to get my hands on for this test. This shouldn't have a noticeable impact on performance but be aware the 4GB models cost around $110 and at that price are a rather poor value.
First up we have Counter-Strike: Global Offensive. Those who are more serious about the game often play using low quality settings and here the GT 1030 and RX 550 deliver a similar experience, though the GT 1030 was slightly better overall.
Both maintained over 100 fps at all times and allowed for the average frame rate to exceed 200 fps. However, I'm told serious gamers like 300 plus fps in this title, but you know, beggars can't be choosers and for the price these results are very acceptable in this title.
If you prefer playing with eye candy enabled, the very high quality preset still allows for over 160 fps on average and here both the RX 550 and GT 1030 deliver the exact same experience.
Overwatch is another super popular first person shooter and it can be much more visually demanding than CS:GO. Even so using the low quality preset the game was quite playable using either GPU, though the GT 1030 was a whisker faster. Here we see the 1% low result was 8% higher using the Nvidia GPU.
That said, increasing the quality setting to the high preset things swing heavily in favor of the RX 550. Before anyone jumps to conclusions and claims this is down to the 4GB VRAM buffer of the RX 550 card used, please note at 1080p using the high quality preset the game never allocates more than 940MB worth of assets at any given time in our 60 second benchmark pass. So the fact that the RX 550 has 4GB of VRAM is of no consequence here.
In the end, for those that like to turn things up a bit higher will be better off with the RX 550 in Overwatch as it allowed for 23% better performance using the high quality preset.
Rocket League is one of the most entertaining and easy to run games available right now and so its massive popularity is unsurprising. If you turn everything down using the high performance preset the game spits out on average 245 fps with the RX 550 and 201 fps with the GT 1030. I'm not sure if you really need that many frames in this title but either way the Radeon GPU was 22% faster.
Then when you max the game out using the high quality preset with MLAA enabled the RX 550 was 24% faster with an average of 77 fps which I thought was quite good at 1080p.
- Ad-free TechSpot experience while supporting our work
- NEW Get a free copy of Superhot: Mind Control Delete, courtesy of GOG
- Our promise: All reader contributions will go toward funding more content
- That means: More tech features, more benchmarks and analysis